auroflow commented on code in PR #28797: URL: https://github.com/apache/flink/pull/28797#discussion_r3652207492
########## flink-python/pyflink/dataframe/convert.py: ########## @@ -0,0 +1,263 @@ +################################################################################ +# Licensed to the Apache Software Foundation (ASF) under one +# or more contributor license agreements. See the NOTICE file +# distributed with this work for additional information +# regarding copyright ownership. The ASF licenses this file +# to you under the Apache License, Version 2.0 (the +# "License"); you may not use this file except in compliance +# with the License. You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +################################################################################ + +from enum import Enum +from typing import Any, Iterable, List, Mapping, Optional, Sequence, Tuple, Union, cast + +from pyflink.dataframe.context import get_or_create_table_environment +from pyflink.dataframe.dataframe import DataFrame +from pyflink.util.api_stability_decorators import PublicEvolving + +__all__ = ["from_dict", "from_records"] + +_SCALAR_SEQUENCE_TYPES = (str, bytes, bytearray, memoryview) + + +class _RecordType(Enum): + NAMED_TUPLE = "named_tuple" + MAPPING = "mapping" + SEQUENCE = "sequence" + + @classmethod + def from_record(cls, record: Any) -> "_RecordType": + if isinstance(record, tuple) and isinstance( + getattr(record, "_fields", None), tuple + ): + return cls.NAMED_TUPLE + if isinstance(record, Mapping): + return cls.MAPPING + if isinstance(record, Sequence) and not isinstance( + record, _SCALAR_SEQUENCE_TYPES + ): + return cls.SEQUENCE + raise TypeError + + +def _validate_schema(schema: List[str]) -> None: + if not isinstance(schema, list) or any(not isinstance(name, str) for name in schema): + raise TypeError("schema must be a list of strings") + if not schema: + raise ValueError("schema must not be empty") + if any(not name for name in schema): + raise ValueError("schema field names must not be empty") + if len(set(schema)) != len(schema): + raise ValueError("schema field names must be unique") + + +def _validate_record_type( + record: Any, expected_record_type: _RecordType, index: int +) -> None: + try: + record_type = _RecordType.from_record(record) + except TypeError: + record_type = None + + if record_type is expected_record_type: + return + # Treat named tuples as tuples when validating sequence records. + if ( + expected_record_type is _RecordType.SEQUENCE + and record_type is _RecordType.NAMED_TUPLE + ): + return + if expected_record_type is _RecordType.NAMED_TUPLE: + raise TypeError("record at index %d must be a named tuple" % index) + if expected_record_type is _RecordType.MAPPING: + raise TypeError("record at index %d must be a mapping" % index) + if expected_record_type is _RecordType.SEQUENCE: + raise TypeError( + "each record must be a sequence of values, " + "such as a list or tuple; invalid record at index %d" % index + ) + raise AssertionError("unsupported expected record type") + + +def _to_field_mapping( + record: Any, expected_record_type: _RecordType, index: int +) -> Mapping[str, Any]: + _validate_record_type(record, expected_record_type, index) + if expected_record_type is _RecordType.NAMED_TUPLE: + fields = cast(Tuple[str, ...], getattr(record, "_fields")) + return dict(zip(fields, record)) + if expected_record_type is _RecordType.MAPPING: + return cast(Mapping[str, Any], record) + raise TypeError("sequence records do not have named fields") + + +def _from_rows(rows: Iterable[Sequence[Any]], columns: Sequence[str]) -> DataFrame: + table = get_or_create_table_environment().from_elements(rows, list(columns)) + return DataFrame(table) + + +@PublicEvolving() +def from_records( + data: Sequence[Union[Sequence[Any], Mapping[str, Any]]], + schema: Optional[List[str]] = None, +) -> DataFrame: + """ + Create a DataFrame from row-oriented records. + + For mapping and named tuple records with an explicit ``schema``, every record must contain all + schema fields; other fields are ignored. When ``schema`` is omitted, the keys or fields from + the first record are used as the schema and every record must have exactly those fields. + + For other sequence records, every record must have the same number of values. A ``schema`` is + required to provide the field names. + + Field types are inferred from the record values. + + :param data: Non-empty sequence of mapping or sequence records. + :param schema: Optional non-empty list of field names. + :return: A DataFrame containing the records. + :raises TypeError: If a record or schema has an invalid type. + :raises ValueError: If data or schema is empty, schema field names are invalid, a required + schema is omitted, a required field is absent, inferred record fields differ, or record + widths differ. + + Example:: + + >>> import pyflink.dataframe as pf + >>> users = pf.from_records([ + ... {"id": 1, "name": "Alice"}, + ... {"id": 2, "name": "Bob"}, + ... ]) + >>> users = pf.from_records( + ... [(1, "Alice"), (2, "Bob")], schema=["id", "name"] + ... ) + >>> from typing import NamedTuple + >>> class User(NamedTuple): + ... id: int + ... name: str + >>> users = pf.from_records([User(1, "Alice"), User(2, "Bob")]) + >>> selected_users = pf.from_records( + ... [User(1, "Alice")], schema=["name", "id"] + ... ) + + .. versionadded:: 2.4.0 + """ + if not isinstance(data, Sequence) or isinstance(data, _SCALAR_SEQUENCE_TYPES): + raise TypeError( + "data must be a sequence of records, such as a list or tuple" + ) + if not data: + raise ValueError("data must not be empty") + + first_record = data[0] + rows: Iterable[Sequence[Any]] + try: + expected_record_type = _RecordType.from_record(first_record) + except TypeError: + raise TypeError( Review Comment: Agreed, this eliminates the need to pass indexes. I refactored this part to use exception chaining. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
